Conventionally, noise is superimposed on a sensor signal for controlling an engine. Thus, a filter for removing the noise is required for acquiring an accurate sensor signal. However, when a filtering process is performed for removing noise, a delay time due to the filtering process (hereafter, a filer delay time) is generated. JP-A-2008-169728 (corresponding to US 2008/0167793 A and referred to as a patent document No. 1) discloses a method of compensating the filter delay time.
In the method of compensating the filter delay time disclosed in the patent document No. 1, all data passing through a digital filter in a predetermined section is stored in a random access memory (RAM). Then, the stored data is processed with another digital filter from a last stored data to a first stored data in order. Accordingly, a digital filter delay time can be compensated.
In a conventional internal combustion engine for a vehicle, an engine control depending on a crank angle is performed and it is required for acquiring an accurate sensor signal synchronized with the crank angle. However, because a filtering process for removing noise is required, a filer delay time is generated. JP-A-2005-220796 (corresponding to US 2005/0166665 A and referred to as a patent document No. 2) discloses a method of compensating the filter delay time.
In the method disclosed in the patent document No. 2, a sensor signal and a counter value of the crank angle at the point are detected, and the sensor signal and the counter value are stored in a memory so as to correspond to each other. Accordingly, a sensor signal corresponding to any crank angle can be acquired.
However, the methods disclosed in the patent document No. 1 and the patent document No. 2 have the following difficulties. In the method disclosed in the patent document No. 1, a huge amount of data passing through the digital filter in the predetermined section is once stored in the RAM and the filter delay is compensated by processing the stored data with the digital filter from the last stored data to the first stored data in order. Because the huge amount of data has to be stored in the RAM, a storage capacity of the RAM is huge and a cost increases.
In the method disclosed in the patent document No. 2, a sensor signal is AD converted at predetermined intervals (e.g., 10 μsec). Thus, a timing of a desired crank angle is less likely to correspond to a timing of the AD conversion, and an accurate data synchronized with the crank angle may not be acquired.